Environ. Sci. Technol. 2001, 35, 2817-2822
The Leaching Behavior of Cement Stabilized Air Pollution Control Residues: A Comparison of Field and Laboratory Investigations ISABEL BAUR,* CHRISTIAN LUDWIG,‡ AND C. ANNETTE JOHNSON Swiss Federal Institute for Environmental Science and Technology (EAWAG), Box 611, CH-8600 Du ¨ bendorf
The factors controlling leachate composition of cement stabilized air pollution control (APC) residues (41% APC residues, 22% cement, 3% Na2CO3, and 32% water, w/w) have been investigated both in the laboratory and in a pilot landfill. Batch leaching and tank leaching tests were carried out in the laboratory in order to determine solubility controlling phases and diffusion controlled species. The major species Ca, SO4, Al, and Si could be partially modeled by assuming calcium silicate hydrate (C-S-H), portlandite, and ettringite to be the solubility controlling phases both in field and laboratory. There were obviously additional minerals that could not be taken into account in calculations because of the lack of data. The determined effective diffusion coefficients (De) for Na and K (2.18e-12 and 5.43e-12 m2s-1) were used to model field concentrations. Agreement with field data was good. Heavy metal concentrations were in the range of 10-8 mol dm-1 (Cd, Co, Cu, Mn, Ni) to 10-6 mol dm-1 (Mo, Pb, W, Zn) in all experiments and often lower in the field leachate than expected from batch experiments. In laboratory experiments, the solubility of Mo and W was most probably controlled by their calcium metalates, Cu by CuO, Ni by Ni(OH)2, and Zn probably by a Zn containing C-S-H phase. In the field, diffusion seems to control Mo and W leachability, with calculated De values of 3.49e-14 and 1.35e-15 m2 s-1.
Introduction Incineration of municipal solid waste produces about 60 000 tons of air pollution control (APC) residues in Switzerland every year. This waste is hazardous because of its high heavy metal content. The use of cement to stabilize hazardous wastes is common practice, and in Switzerland a significant proportion of APC residues from municipal solid waste incineration (MSWI) is stabilized with cement prior to landfilling. The APC residues are washed in order to remove salts, dried, and mixed with cement in a ratio of approximately 2:1. During hydration, heavy metal species can dissolve, adsorb to the matrix, or be incorporated in the newly formed (cement) minerals. The latter case has been shown to be important for e.g. Zn (1) and As (2). The evaluation of the leaching behavior is an important aspect in waste management for the long-term protection * Corresponding author phone +41 1 823 5461; fax: +41 1 823 5210; e-mail:
[email protected]. ‡ Paul Scherrer Institute (PSI), 5232 Villigen PSI, Schlieren, Switzerland. 10.1021/es000243r CCC: $20.00 Published on Web 05/25/2001
2001 American Chemical Society
of the environment, whether materials to be landfilled or reused. There are a variety of tools used to predict both the present day and long-term leaching behavior that utilize data from laboratory and field experiments. Whereas field experiments are largely restricted to the monitoring of pilot studies, laboratory tests employ a number of different approaches (e.g. ref 3): Equilibrium experiments using ground samples provide information on mechanisms that control species solubility. These include the precipitation of discrete phases, sorption, and complexation. Dynamic leaching tests directly try to interpret leaching processes, including diffusion, dissolution, and surface wash-off (4). Column tests are used to imitate field conditions, where both dissolution/precipitation as well as diffusion processes under conditions with a low liquid to solid (L/S) ratio are important (5). Field studies concentrate on leachate composition with emphasis on hydrology and dynamic processes (6-9). A comparison of results from field investigations with those from laboratory experiments can facilitate the identification of the processes that control leachate composition. The comparability of laboratory and field leachates is based on the assumption that not only are the conditions (such as pH) and the processes that control contaminant mobility the same but that the time frame of the leaching process is the same too. To date only a few studies have compared field and laboratory data. Four studies on coal fly ash made important contributions (10, 11, 5, 12). Dodd et al. (10) were not able to explain field leachate composition with laboratory leaching tests. They pointed out that laboratory leaching tests could supply important information on the leachate characteristics such as pH-range and conductivity but could rarely simulate the complexity of the field system. Fruchter et al. (11) identified potential solubility controlling phases for several major and minor species. From laboratory tests the authors concluded that equilibrium times were in order of days or less, thus different water residence time scales in field and laboratory experiments would not affect the results for granular materials such as coal fly ash. Hjelmar et al. (5) found that laboratory and field results were comparable at the same L/S ratio. However, they stressed that kinetic processes as pozzolanic and oxidizing reactions occurring in the untreated residues after the first contact with percolating water may cause strongly differing results. Meij and te Winkel (12), who kept the L/S ratio in the field and laboratory experiments identical, also attributed observed differences in leachate compositions to kinetic factors. There appears to be a complex interplay between processes that can achieve equilibrium and those that are kinetically controlled. While the previous mentioned studies were on unsolidified waste, Stegemann et al. (13) performed a study on solidified electric arc furnace slag. They pointed out that in the case of a monolithic waste, where the interaction between water and waste is restricted, diffusion controlled leaching becomes important. The aim of this paper is to illustrate and compare the dominant processes that control leachate composition in the laboratory and the field. The concentration of leachate species will be explained in terms of dissolution/precipitation or diffusion processes and discrepancies between field and laboratory data are discussed. A batch leaching test has been used to achieve quasiequilibrium between the sample and the leachate. The resulting data can be compared to thermodynamic models to determine which solid phases are likely to control solubility of leachate species. A tank leaching test similar to ANS-16.1 (14) has been performed to derive effective diffusion coefVOL. 35, NO. 13, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2817
ficients for the modeling of diffusion-controlled species concentrations. The hydrology and partly the leachate composition of our field site, the lysimeter Teuftal (Kanton Bern, Switzerland), has been investigated by Ludwig et al. (9). It could be shown that the leachate is of almost constant composition and becomes diluted during rain events as a consequence of preferential flow. The same leachate behavior has been reported for a bottom ash landfill (6). The mean water residence time of around 2 months in the lysimeter is assumed to be long enough to allow the water to interact and equilibrate with at least some species of the solidified waste (9).
Experimental Section Field. The field lysimeter was constructed in the winter 1990/ 1991 next to the landfill Teuftal in Frauenkappelen, Switzerland. The investigated lysimeter compartment (15-17, 9) contained cement-stabilized APC residues in the form of blocks with an edge length of 0.5 m. They were prepared with 43% APC, 22% cement, 3% sodium carbonate, and 32% water (w/w). The lysimeter was approximately 1.5 m deep, had a surface area of 16 m2, and was covered with geotextile, gravel (0.8 m) and humus (0.3 m) layers. It was dismantled in December 1997. Leachate sampling is described in ref 9. Laboratory. During the dismantling of the field lysimeter in December 1997, drill cores (0.1 m diameter, 0.3 m length) were taken from the blocks, packed in hermetically sealed plastic bags with soda lime to prevent CO2 uptake, and kept at -20 °C until used in the laboratory leaching experiments. All handling of the sample material in the laboratory was undertaken in a CO2-free atmosphere. Water was taken from a 17MΩ ultrapure system (Barnstead Nanopure), with a 0.2µm inline filter. Prior to use, the ultrapure water was boiled and cooled to 20 °C under Argon. Polyethylene bottles were leached with acid (∼0.1 M diluted from concentrated HNO3) for at least 24 h and rinsed with ultrapure water. The same procedure, but with a more diluted HNO3-solution (∼pH 4), was used to leach and rinse 0.45 µm Nylon filters (Whatmann). Batch Leaching Test. The material was dried and ground to 0.990, with the exception of Cu and Mo where r2 was 0.983 and 0.977, respectively). Nevertheless, it is important to mention that the chosen testing times result in constant leachate concentrations and a linear CFR vs square root of time-relationship can be both the result of a saturated solution or diffusive release. This complicates the distinction between solubility controlled and diffusion controlled release. No De values were calculated for Cd, Co, and Mn because their concentration in the leachate was below detection limit (
